Evolution of random networks

Abstract

We investigate extremal dynamics on random networks. In the quenched case, after a transient time the dynamics is localized in the largest cluster. The activity in the largest cluster is nonergodic, with hot spots of activity typically centered around nodes with a high coordination number. The nonergodicity of the activity opens for models of evolving networks, which can self-organize into fractal geometries. © 1998 The American Physical Society.